Ask ten superintendents what a weekly work plan is and you'll get ten versions of the same answer: it's the list of what the crews are doing this week. True enough. But that answer misses why the weekly work plan matters, and why the good jobs run on it while the bad ones just fill it out on Friday and forget it by Tuesday. The weekly work plan is the sharp end of the Last Planner System. It's where all the upstream planning either pays off or falls apart, and understanding that relationship is the difference between a plan you follow and a plan you file.
I've watched crews blow through beautiful pull plans because nobody connected the phase schedule to what the electrician was actually going to touch on Wednesday. The weekly work plan is that connection. Let me walk through where it sits, what it's supposed to do, and the specific ways it goes sideways on real jobs.
Where the Weekly Plan Sits in Last Planner
The Last Planner System is really a set of nested planning horizons, each one tighter and more reliable than the one above it. From the top down:
- Master schedule — the milestone-level plan, the CPM your project team lives and dies by with the owner. This says should, in the broadest sense.
- Phase (pull) planning — the trades sit in a room and work backward from a milestone, sticky notes on the wall, handoff by handoff. This is where sequence logic actually gets built by the people doing the work.
- Look-ahead planning — a rolling window, usually six weeks, where you run the make-ready process: identify constraints and knock them down before the work is due.
- Weekly work planning — this week's commitments. Only work that's actually ready.
- Daily coordination — the huddle that keeps the week honest.
Notice the language shifts as you move down. The master schedule says work should happen. The look-ahead asks whether it can happen — is it free of constraints? The weekly work plan is where a foreman says it will happen. That should-can-will progression is the entire point, and the weekly plan is where "will" gets spoken out loud in front of the other trades.
Reliable Promising: The Whole Ballgame
Here's the idea that separates Last Planner from ordinary scheduling. A commitment on the weekly plan is a promise, not a prediction. The general contractor doesn't assign the work; the foreman who owns the crew commits to it. That's why they're called the last planners — they're the last people between the plan and the actual installed work, and they know things the office schedule never will.
The rule that makes promising reliable is brutally simple: you don't commit to work that isn't ready. If the framing inspection hasn't cleared, the drywall foreman doesn't put "hang rock, C-wing" on the plan, no matter how badly the master schedule wants it. A commitment made on unready work isn't optimism, it's a lie the whole team agrees to tell itself, and it poisons every downstream trade that was counting on that handoff.
The hard part is cultural, not technical. Foremen are conditioned to say yes. Saying "no, that's not ready, I'm not committing" in a room full of superintendents takes some spine, and it only works if the GC has made it clear that an honest "no" is more valuable than an optimistic "yes" that busts on Thursday. Protect that. The first time you punish a foreman for a truthful "not ready," you've taught everyone to go back to lying.
Make-Ready: Why the Work Is Actually Ready
Weekly commitments are only as good as the make-ready work that fed them. In the look-ahead window you're screening every upcoming activity against the standard constraint categories. The classic list, and I'd run through all of it every week:
- Prerequisite work — is the preceding trade actually finished and inspected, not just "mostly done"?
- Design and RFIs — any open question that could change what you build?
- Materials and submittals — approved, ordered, and on site or with a firm delivery date inside the window?
- Labor — the right crew size and the right skills available that week?
- Equipment — lift, hoist, crane time booked?
- Permits and inspections — scheduled, with the inspector's lead time accounted for?
- Access and conditions — is the area clear, is it dry, is temporary power there, can two trades physically fit?
A constraint you find three weeks out is a phone call. The same constraint discovered the morning the crew shows up is a lost day and a pissed-off subcontractor. That's the entire economic argument for make-ready: constraints get cheaper to fix the earlier you find them. The weekly work plan should be nothing but the residue of a make-ready process that's already cleared the path.
PPC: Measuring Whether Your Word Is Good
Percent Plan Complete is the one metric that keeps the whole system honest. It's not complicated: at the end of the week, count the commitments that were 100% complete, divide by the total number committed. Committed eight, finished six, your PPC is 75%. No partial credit — a task that's 90% done counts as a zero, because a handoff that's 90% ready isn't a handoff.
Most crews new to this land somewhere in the 50s at first, and that's fine; it's a starting line, not a verdict. Mature teams run consistently in the 80s. Chasing 100% is usually a warning sign, not a triumph — it often means foremen are lowballing their commitments to protect the number. If your PPC is pinned at 100, your plan is probably too soft.
The trap is treating PPC as a report card. It isn't. It's a thermometer. A dropping PPC tells you your make-ready process is leaking, and that's useful information you'd want to have.
Variance Analysis: The Part Everyone Skips
PPC without variance analysis is just a number on a whiteboard. The real learning is in the tasks that didn't complete, and specifically in why. Every missed commitment gets a reason code, and you want to categorize honestly:
- Prerequisite work wasn't done
- Materials or information late
- Change in priorities / crew pulled to another area
- Bad estimate — the work just took longer than the foreman thought
- Weather or conditions
- Design change / RFI
Do this for a month and patterns jump out. If half your misses trace to "prerequisite work not done," your problem isn't the weekly plan, it's the trade upstream, and the conversation moves to phase planning. If they trace to "materials late," you've got a procurement or submittal problem the office needs to own. The point of variance analysis is to fix the system, not to find someone to blame for Thursday. The moment it turns into a blame exercise, the honest reason codes disappear and you're back to guessing.
The Weekly Meeting That Ties It Together
The weekly Last Planner meeting has a rhythm, and it's worth running the same way every time so people know what's coming. Keep it tight, keep the foremen talking, and don't let it turn into the superintendent lecturing:
- Score last week. Walk the prior plan, mark each commitment done or not, calculate PPC in front of everyone.
- Analyze the misses. For every incomplete task, land on a reason code. Two minutes each, no defending, just the truth.
- Refresh the look-ahead. Update constraint status on the six-week window. Who owns each open constraint, and when does it clear?
- Make new commitments. Foremen — not the GC — commit to next week's ready work, trade by trade, out loud, with the handoffs visible so the plumber hears what the framer just promised.
That last part is where trade flow becomes real. When the mechanical foreman commits to finishing overhead in room 210 by Wednesday and the drywall foreman is sitting right there planning to close that ceiling Thursday, the sequence is coordinated by the two people who actually control it. That conversation is worth more than any Gantt chart.
Where the Software Earns Its Keep
You can run Last Planner with sticky notes and a spreadsheet, and plenty of great jobs have. But the bookkeeping gets heavy fast — tracking constraints across a rolling window, rolling incomplete work forward, calculating PPC, tagging variances week over week, and keeping every trade looking at the same current plan. That's the honest role for a tool. Good look-ahead scheduling software carries the administrative load so the team can spend the meeting talking about the work instead of arguing about whose spreadsheet is current.
This is where a visual, location-based platform like LookAheadWall fits naturally: the weekly work plan lives in the same place as the trade-flow sequences, so a foreman's commitment is tied to the actual location and the handoff that follows it, and the schedule the subs pull up on their phones is the same one that was committed to in the meeting. The tool doesn't make the promises — the foremen still do that — but it keeps the plan visible, current, and shared, which is exactly where paper-based systems tend to break down on a busy job.
Pick a tool because it supports the method, not the other way around. If a platform can't track constraints across the look-ahead window, can't roll incomplete commitments forward, and can't spit out PPC and variance without you rekeying everything, it's a calendar with a nice logo, not a Last Planner tool.
How to Start Without Blowing It Up
If you're bringing this to a crew that's never run it, don't launch the full apparatus on day one. A few things that tend to work:
- Teach the method before the software. People who understand why they're screening constraints will actually do it; people just filling in an app won't.
- Run make-ready for a couple of weeks before you start scoring PPC. Let the pipeline fill so the first weekly plans are built on ready work.
- Start PPC as a private team number, not something reported up to the owner. It's a learning tool first. Broadcast it too early and the foremen start gaming it.
- Bring the subs in early. Last Planner without the trades committing is just the GC talking to itself.
The weekly work plan isn't a standalone document, and it never was. It's the point where master schedule intent, phase-plan sequence, and make-ready constraint work all converge into a set of promises somebody's actually willing to make. Get that connection right and the plan starts predicting the future instead of describing the past. Get it wrong and you've got a very tidy list of things that didn't happen.